Dependency trees, permutations, and quadratic assignment problem

ABSTRACT

This paper describes and analyzes an estimation of distribution algorithm based on dependency tree models (dtEDA), which can explicitly encode probabilistic models for permutations. dtEDA is tested on deceptive ordering problems and a number of instances of the quadratic assignment problem. The performance of dtEDA is compared to that of the standard genetic algorithm with the partially matched crossover (PMX) and the linear order crossover (LOX). In the quadratic assignment problem, the robust tabu search is also included in the comparison.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.6 Learning

Additional Classification:
  G. Mathematics of Computing
  G.1 NUMERICAL ANALYSIS
      G.1.6 Optimization

  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.8 Problem Solving, Control Methods, and Search

General Terms:
Algorithms

Keywords:
dependency tree, estimation of distribution algorithm, linkage learning, permutation domain, quadratic assignment problem